Cell migration is a critical feature of normal development and disease. For >25 years, we have been developing an in vivo model in which we combine sophisticated genetic and optogenetic manipulations with quantitative live imaging analyses to probe molecular mechanisms underlying collective cell migration: the border cells in the Drosophila ovary. When we began, we did not know a single gene that was required, and live imaging was impossible. Now many molecular pathways are known. We were the first to show that the small GTPase Rac controls protrusion and migration in vivo. We went on to use a photo-activatable form of Rac to show that activation in a single cell is sufficient to steer the entire cluster. Now Rac and its relatives Rho and Cdc42 are well known as key nodes in the signaling and cytoskeletal pathways that govern cell polarity and migration. Recently it has become clear that tumor cells disseminate in groups that resemble border cell clusters in several key aspects. So it is more relevant and interesting than ever to investigate the underlying mechanisms. Here we propose to continue our longstanding pattern of technical and conceptual innovation to address a key outstanding question. How do the molecular mechanisms of cell motility derived from studies of single cells migrating unobstructed on glass applies to the more diverse morphologies and behaviors of cells traveling in groups in vivo through 3D, cell-rich terrains. In cells migrating individually in vitro, mutually inhibitory interactions between Rac and Rho set up distinct protruding and contractile domains. It is unclear how this model applies to groups of cells moving in vivo. In Aim 1 we will address the following key open questions: In collectively moving cells, do Rac and Rho inhibit one another? Do they do so cell autonomously, non- cell-autonomously, or both? Is Rac only required in the lead cell for protrusion? Is Rho specifically required in following cells? Individually migrating cells can also switch from a Rac-dominated/protrusive mode of migration to a Rho-dominated, contractile mode. Do collectively moving cells exhibit comparable plasticity? In Aim 2, we propose to decipher the roles of apical/basal polarity complexes in collective chemotaxis. Individually migrating cells lack apical/basal polarity but collectively migrating cells require it. We propose to tease apart the autonomous and non-autonomous contributions of apical and basolateral protein complexes to the coordination of collective cell motility. Finally, in Aim 3 we propose to unify seemingly disparate observations into a common conceptual framework to test the hypothesis that multiple downstream effectors of Cdc42, all of which contribute to border cell migration, form an integrated network. These studies will lead to a more precise and comprehensive understanding of the intracellular signaling networks that provide cells with 3D coordinates. Our work will produce a paradigm for us to test in collaboration with our colleagues who study collective cell dissemination in metastasis.